Glycerol dialkyl glycerol tetraethers (GDGTs), including both the archaeal isoprenoid GDGTs (isoGDGTs) and the bacterial branched GDGTs (brGDGTs), have been used in paleoclimate studies to reconstruct temperature in marine and terrestrial archives. However, GDGTs are present in many different types of environments, with relative abundances that strongly depend on the depositional setting. This suggests that GDGT distributions can be used more broadly to infer paleoenvironments in the geological past. In this study, we analyzed 1153 samples from a variety of modern sedimentary settings for both isoGDGT and brGDGTs. We used machine learning on the GDGT relative abundances from this dataset to relate the lipid distributions to the physical and chemical characteristics of the depositional settings. We observe a robust relationship between the depositional environment and the lipid distribution profiles of our samples. This dataset was used to train and test the Branched and Isoprenoid GDGT Machine learning Classification algorithm (BIGMaC), which identifies the environment a sample comes from based on the distribution of GDGTs with high accuracy. We tested the model on the sedimentary record from the Giraffe kimberlite pipe, an Eocene maar in subantarctic Canada, and found that the BIGMaC reconstruction agrees with independent stratigraphic information, provides new information about the paleoenvironment of this site, and helps improve paleotemperature reconstruction. In cases where paleoenvironments are unknown or are changing, BIGMaC can be applied in concert with other proxies to generate more refined paleoclimatic records.

Ancient lake deposits in the Mojave Desert indicate that the water cycle in this currently dry place was radically different under past climates. Here we revisit a 700 m core drilled 55 years ago from Searles Valley, California, that recovered evidence for a lacustrine phase during the late Pliocene. We update the paleomagnetic age model and extract new biomarker evidence for climatic conditions from lacustrine deposits (3.373–2.706 Ma). The MBT5Me′ temperature proxy, based on bacterial membrane lipids, detects present-day conditions (21 ± 3 ºC, 1s, n = 2) initially, followed by warmer-than-present conditions (25 ± 3 ºC, n = 17) starting at 3.268 and ending at 2.734 Ma. This is supported by salinity indicators from bacterial and archaeal biomarkers that reveal lake salinity increased after 3.268 Ma. The δ13C values of plant waxes (-30.7 ± 1.4‰, n = 28) are consistent with local C3 taxa, likely expanded conifer woodlands during the pluvial with less C4 than the Pleistocene. dD values (-174 ± 5‰, n = 25) of plant waxes indicate precipitation dD values (‑89 ± 5‰, n = 25) in the late Pliocene are within the same range as the late Pleistocene precipitation dD. Microbial biomarkers identify a deep, freshwater lake and a cooling that corresponds to the onset of major Northern Hemisphere glaciation at marine isotope stage MIS M2. A more saline lake persisted for ~0.6 Ma across the subsequent warmth of the late Pliocene before the lake desiccated at the Pleistocene intensification of Northern Hemisphere Glaciation.

Great Salt Lake (UT) is a hypersaline terminal lake in the US Great Basin, and the remnant of the late glacial Lake Bonneville. Holocene hydroclimate variations cannot be interpreted from the shoreline record, but instead can be investigated by proxies archived in the sediments. GLAD1-GSL00-1B was cored in 2000 and recently dated by radiocarbon for the Holocene section with the top 11 m representing ~7 ka to present. Sediment samples every 30 cm (~220 years) were studied for the full suite of microbial membrane lipids, including those responsive to temperature and salinity. The ACE index detects the increase in lipids of halophilic archaea, relative to generalists, as salinity increases. We find Holocene ACE values ranged from 81-98, which suggests persistent hypersalinity with <50 g/L variability across 7.2 kyr. The temperature proxy, MBTʹ5Me,­ yields values similar to modern mean annual air temperature for months above freezing (MAF = 15.7°C) over the last 5.5 kyr. Several GDGT metrics show a step shift at 5.5 ka before which temperature estimates are unreliable due to the shift in lake ecology and likely shallow depth. The step change in lake conditions at 5.5 ka and additional variations within the late Holocene are compared to regional climate records. We find evidence for a dry mid-Holocene in GSL, corroborating other records.